About AutoMark
AutoMark is a smart mobile application designed to help educators automate the process of grading student scripts. Whether you're handling a few papers or a large stack, AutoMark makes grading faster and easier by using advanced technologies like Optical Character Recognition (OCR) and Artificial Intelligence (AI).
Teachers can upload student scripts, provide marking guides, and choose between Auto Mark or AI Mark to instantly generate accurate results. Built with Flutter and Firebase, AutoMark delivers a smooth, secure, and responsive experience.
For users who need even more power, AutoMark offers a Premium plan with features such as Bulk Marking, an Advanced Analytics Dashboard, Detailed PDF Reports, and Long-Term Script Storage. Payments are securely handled through MTN Mobile Money (MoMo), making it simple and accessible for educators.
AutoMark saves time, reduces human error, and provides clear feedback — making grading smarter, not harder.
Meet the Team

Lyazi Patrick
Lead Developer & Integrator
Coordinated development and integrated numerous screens into the project.
Ensured seamless navigation and user experience across the app and website.

Ssekidde Jovan
UI Designer & Web Developer
Designed and developed application screens and website.
Created visually appealing and user-friendly interfaces for both web and mobile.

Wanswa Drake
OCR & AI Specialist
Integrated OCR and improved model accuracy.
Enabled the app to accurately read and grade handwritten scripts using advanced OCR technology.

Milla Samantha
UI/UX Engineer
Built the UI/UX for mobile and web.
Crafted intuitive and attractive user experiences for both the app and website.

Adrian Katongole
Authentication & Storage Lead
Set up authentication and real-time storage.
Ensured secure user login and reliable data storage for all users.
How the ML & OCR Works
We use Google's ML Kit Text Recognition API to scan handwritten or printed text from question papers, marking guides, and answer sheets. The recognized answers are then matched against correct responses stored in the app's Firebase database. This allows for: